Ken Griffin Says AI Is “Almost Certainly” in a Bubble
Ken Griffin says AI is almost certainly in a bubble, warning chip prices could fall, but expects surviving companies to lead for decades.
Agentic Times contributors
Published articles
NVIDIA DSX Platform Targets Power Bottlenecks in AI Factory Buildouts
NVIDIA says its DSX platform helps companies plan power, cooling and computing together as grid delays slow AI factory construction.
NVIDIA Details European Government AI Projects Built on Open Models
NVIDIA detailed European government AI projects using open models to keep sensitive data under local control across six regions.
Ex-Nvidia VC Amulya Vishwanath Says Physical AI Is in Its “LEGO Era”
Former Nvidia marketer Amulya Vishwanath says physical AI is in a “LEGO era,” with robotics building blocks still being developed.
Hackers Drain AI Accounts After OpenCode Console Breach
Hackers breached an older OpenCode system, exposing users’ AI API keys and causing unauthorized charges, and the company says refunds are underway.
Google Releases TorchTPU as Open Source to Challenge Nvidia in AI
Google released TorchTPU on Tuesday, a day after Amazon launched TorchNeuron, giving AI developers new ways to run PyTorch apps without Nvidia chips.
Netflix Documentary on Rogue OpenAI Agents Draws Pushback From Hugging Face
Hugging Face CEO Clem Delangue says Netflix’s documentary on the OpenAI agent hack exaggerates events and leaves out key details about the breach.
Anthropic Drafts AI Safety Bill That Could Block Models Risking 50+ Deaths
Anthropic has drafted a federal bill that would let a court block AI models posing risks of 50+ deaths or $1 billion damage.
Spotify Removes 50 Million AI Tracks From Recommendations in 2026
Spotify removed over 50 million AI-generated tracks from recommendations this year, using listener skips, metadata and detection tools to protect human artists.
IBM Explains How AI Decision Models Handle Routine Work Choices
An IBM article explains how AI decision models pick quick answers with confidence scores and send low-confidence choices to humans for review.
European Grant Applications Rise 57% Since ChatGPT Launch, Data Shows
Research funders say AI-written peer reviews are adding pressure as grant applications rise 57%, while experts stress human oversight remains essential.
IBM Research Says llm-d Cut AI Agent Costs From $300 to $26 in 24 Hours
The open source llm-d tool reached version 1.0 this week, with backers saying it cuts AI inference costs on GPU clusters.